Two numbers, same question, same afternoon, both produced by us. The first comes from the exponential average carried on every row of our market scan. The second comes from a separate breadth calculation that uses a simple average over its own list of names, where the denominator is 1,688 rather than 1,677.
Neither is wrong. They are answers to slightly different questions that happen to be phrased identically, and that is worth understanding before you quote anyone’s market breadth figure, including ours.
What the two averages actually do
A simple moving average adds up the closing prices over the lookback and divides. Every day in the window counts exactly as much as every other day, including the one 199 sessions ago.
An exponential moving average weights recent days more heavily and older days progressively less. It bends toward new prices sooner. That makes it quicker to reflect a change in direction and, for the same reason, quicker to react to noise that turns out to be nothing.
Neither is the better tool. The simple average is smoother and slower, the exponential is faster and jumpier, and the choice is about which failure you would rather have.
Because they sit at different levels on any given day, a stock can be above one and below the other at the same moment. Aggregate that across seventeen hundred names and you get two different percentages.
Three differences, not one
It would be tidy to say the gap is the SMA against the EMA. It is not only that, and the other two matter just as much:
- The indicator. One is exponential, one is simple.
- The universe. The two calculations run over different lists of names, which is why the sample sizes are 1,677 and 1,688 and not one shared figure. Neither denominator is the full scan, because a stock needs roughly 200 sessions of history before it has a 200-day average at all.
- The capture time. One is a live intraday scan, the other a daily row. The market moves in between.
The part that surprised us
The obvious story is that one method reads consistently higher than the other, and you could correct for it. That story is wrong.
On the 200-day window our exponential figure was higher: 65.2% against 64.57%. On the 50-day window, from the same capture, it was lower: 55.8% of 1,708 against 56.29% of 1,732.
The sign flips. There is no constant offset, no conversion factor, and no way to translate a breadth number computed one way into the number you would have got the other way. The only thing that resolves it is knowing which one you are looking at.
We shipped this bug ourselves
This is not a hypothetical. On 16 August 2026 a lesson on this site and a dated market note published by our own analyst gave slightly different figures for the share of stocks above their 200-day average, for the same day.
The first explanation was the obvious one: different denominators. That was wrong, and it took another day to establish why. They were not one measurement counted two ways. They were two measurements. Only the 200-day figure disagreed, because almost every name has a 50-day average and the two methods nearly agree over a short window, which is exactly the pattern you would expect once you know what is going on.
The fix was not to pick a winner. Every snapshot we take now records both figures, each with its own sample size and the name of the indicator that produced it, in one dated file. A number you cannot trace back to the run that produced it is a number you cannot defend.
What to do with this when you read someone else’s number
“Sixty-five percent of stocks are above their 200-day” is not a portable fact. Before it means anything, three things have to be attached to it: which average, over which list of stocks, measured when.
This matters most when a figure is tracked over time. A move from 64.5% to 65.2% looks like the market broadening. If the two readings came from different methods, nothing happened at all.
What these numbers assume
- Both averages are computed on daily closes. The exponential figures use the 200 and 50 period EMAs carried on our scan rows; the simple figures come from a separate breadth calculation over its own universe.
- Every percentage here carries its own sample size, and they differ on purpose. A stock with less than 200 sessions of history is absent from the 200-day figures rather than counted as being below.
- All four figures come from a single capture on 20 August 2026, archived together. An earlier capture the same day read 56.4% for the 50-day EMA where this one reads 55.8%, because the scan re-ran and the market moved. That is the same class of difference the page is about.
Frequently asked questions
What is the difference between an SMA and an EMA?
A simple moving average gives every day in the lookback the same weight. An exponential moving average weights recent days more heavily, so it turns sooner when price changes and lags less. Neither is more correct. The SMA is smoother and slower, the EMA is faster and noisier, and they are built for different jobs. The practical consequence is that they cross price at different moments, so a stock can be above one and below the other at the same instant.
Why did your two measurements of the same thing disagree?
Three reasons at once, and we think naming all three is more useful than pretending there is one. The indicator differs (an exponential average against a simple one). The universe differs, so the denominators are 1,677 and 1,688 rather than one shared number. And the capture time differs: one is a live intraday scan, the other a daily row. On 20 August 2026 those combined to 65.2% against 64.57% for the share of stocks above their 200-day average.
Does the exponential average always read higher?
No, and that is the most useful thing in this measurement. On the 200-day window our exponential figure was higher, 65.2% against 64.57%. On the 50-day window it was lower, 55.8% against 56.29%. The sign flips between the two windows, so there is no constant offset and no way to convert a number computed one way into a number computed the other. You have to know which one you are reading.
Does a 0.63 point difference actually matter?
For deciding whether the market is broadly rising, no. Both numbers say roughly two thirds of tracked stocks are above their long average and both would lead to the same reading of conditions. It matters when a figure is quoted as a fact, compared against another source, or tracked over time, because then a gap of this size can be mistaken for a change in the market when it is only a change in the method.
Which of your two numbers should be treated as the real one?
Whichever one names its method. We now archive both in the same dated file so a published figure can be traced back to the run and the indicator that produced it, rather than compared from memory. A percentage without its sample size and its definition is not evidence, and that is true of any source quoting market breadth, not only this one.
Related
The same habit of checking a number against its base rate runs through what the covered call regime check misses and what counts as a good Sharpe ratio. For how rarely a fixed indicator threshold actually fires, see what RSI overbought and oversold actually mean. More lessons are on the learn page, and the about page explains who writes them and why.